This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [RFC] ChangeLog generation script
- From: Rical Jasan <rj at 2c3t dot io>
- To: Zack Weinberg <zackw at panix dot com>, Paul Eggert <eggert at cs dot ucla dot edu>
- Cc: Florian Weimer <fweimer at redhat dot com>, Andreas Schwab <schwab at suse dot de>, Szabolcs Nagy <Szabolcs dot Nagy at arm dot com>, nd <nd at arm dot com>, Joseph Myers <joseph at codesourcery dot com>, Siddhesh Poyarekar <siddhesh at gotplt dot org>, GNU C Library <libc-alpha at sourceware dot org>, Carlos O'Donell <carlos at redhat dot com>, Adhemerval Zanella <adhemerval dot zanella at linaro dot org>
- Date: Sat, 24 Nov 2018 12:19:19 -0800
- Subject: Re: [RFC] ChangeLog generation script
- References: <8e00de8e-c9f3-92a5-0382-8da54742e7fb@gotplt.org> <alpine.DEB.2.21.1811192249530.10031@digraph.polyomino.org.uk> <c530622e-80ca-1ea2-eb78-512a13f17ad0@gotplt.org> <alpine.DEB.2.21.1811201648140.27258@digraph.polyomino.org.uk> <mvmy39nlm3d.fsf@suse.de> <alpine.DEB.2.21.1811201705280.27258@digraph.polyomino.org.uk> <mvmin0qlrsq.fsf@suse.de> <7a05db3d-e982-1080-363f-cd3536275779@arm.com> <mvm36rulovy.fsf@suse.de> <5ceb52dc-0bd4-0b8a-eaa0-857525abc36c@arm.com> <mvmy39mka1s.fsf@suse.de> <87in0qaa0t.fsf@oldenburg.str.redhat.com> <CAKCAbMiRyHJxC_anrPvaUkTJVLda_BquduBxr9uJnKB=QzG+Qg@mail.gmail.com> <4874e41b-108f-e1bb-447b-72130ef6798d@cs.ucla.edu> <CAKCAbMjTwxwMvLYYCDH4ta4sFNsGsYb=2X0qgy0Jafsnd2Tf6g@mail.gmail.com>
On 11/24/2018 07:43 AM, Zack Weinberg wrote:
> On Sat, Nov 24, 2018 at 3:10 AM Paul Eggert <eggert@cs.ucla.edu> wrote:
>> Zack Weinberg wrote:
>>> I find *writing*
>>> changelog entries by hand to be a valuable discipline. It forces me
>>> to read every line of the patch I'm about to submit, and I've caught
>>> mistakes dozens of times because of it.
>>
>> That's fine, and you can continue doing that even after we have a script. Other
>> "stop and think twice" techniques work better for me, in the sense that they
>> find mistakes more efficiently than ChangeLog-writing does. (Reading the diff is
>> one of these techniques.) The goal is developing and maintaining the best
>> software we can with the limited resources we have.
>
> To be clear, I wasn't saying that everyone should do it my way; I was
> only presenting an example of why someone who finds very little use
> for the historical information in the ChangeLog files might still not
> mind writing entries by hand (and therefore wants to concentrate on
> other ways in which ChangeLog maintenance is a pain).
I second the sentiment that writing entity-level ChangeLog entries is
nice due to the forced review it imposes, but there are types of changes
where it can be tedious and not particularly useful. If it could be
automated, I think the overall situation might be better for everybody.
Making the script available to run before committing, for review, would
be nice, and probably help flesh out issues with the script.
Also, for the record, I've only ever found ChangeLogs useful when
delving through pre-git history, trying to find when a given entity was
first introduced, for example, and even then the results can be hit or miss.
Rical